Management method for dynamically optimizing scenic spot people flow distribution

Abstract

The invention relates to the technical field of scenic region people flow management, in particular to a management method for dynamically optimizing scenic region people flow distribution, which comprises the following steps of: acquiring and analyzing tourist density data in real time, constructing a tourist density subsequence, calculating a fluctuation characteristic value, evaluating a flow deviation degree, and combining trend term decomposition and sliding window analysis. And calculating a flow anomaly index of each region, and realizing a global optimization decision through a neighborhood region set. According to the method, the dynamic optimization capability of the scenic spot people flow distribution can be improved, the regulation and control efficiency in a peak period or an emergency is enhanced, the tourist experience is improved, and the operation safety and stability of the scenic spot are guaranteed.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent scenic area management, and specifically relates to a management method for dynamically optimizing the distribution of tourist flows in a scenic area. Background Art

[0002] With the rapid development of the tourism industry, managing visitor flow within scenic areas has become a crucial issue for ensuring both a safe and enjoyable travel experience. Existing scenic area crowd flow management systems primarily rely on density analysis using information collected by video surveillance equipment, providing real-time distribution information to aid decision-making. However, existing technologies still have limitations when it comes to dynamically optimizing crowd flow distribution within scenic areas, and the ability to intelligently regulate and guide traffic flow needs to be improved.

[0003] A search revealed a "System and Method for Analyzing Tourist Distribution in a Scenic Area" with publication number CN112668375B, dated July 7, 2023. This patent captures image information from video recording devices distributed throughout the scenic area and processes the images using a combination of LBP texture analysis and convolutional neural networks to obtain visitor density data. This data is transmitted to a cloud platform and displayed on a scenic area map, allowing visitors to access and select itineraries via mobile devices. However, this technical solution focuses on the collection and display of visitor density data and is limited in its ability to dynamically optimize the distribution of visitor flows within the scenic area. Specifically, it lacks sufficient proactive intervention mechanisms based on real-time data, such as dynamically adjusting the openness of various areas within the scenic area or specific strategies for diverting visitor traffic, potentially limiting its effectiveness in alleviating peak-hour congestion. Furthermore, the system's effectiveness is highly dependent on the autonomous choices of visitors, potentially impacting its ability to respond quickly and achieve global optimization when visitor numbers are high or when emergencies occur.

[0004] Therefore, based on the above analysis, it can be seen that the existing technology has the following shortcomings in the management of scenic spot crowd distribution: First, the dynamic optimization and active intervention mechanisms are relatively weak, which affects the effect of flexibly adjusting the allocation of scenic spot resources according to real-time data; second, it is highly dependent on tourist behavior, and there are certain challenges in achieving global crowd control during peak periods or emergency situations; third, it fails to fully combine intelligent algorithms and multi-source data for comprehensive analysis, which restricts the further improvement of management efficiency and accuracy.

[0005] The above issues demonstrate that existing scenic spot crowd flow management technologies still have room for improvement in dynamic optimization, proactive intervention, and intelligent decision support. Therefore, this invention aims to provide a management method for dynamically optimizing scenic spot crowd flow distribution. By incorporating real-time data analysis, dynamic adjustment strategies, and intelligent guidance mechanisms, this method achieves efficient management and optimization of crowd flow distribution within a scenic spot, thereby enhancing the visitor experience and ensuring the safety and stability of scenic spot operations. Summary of the Invention

[0006] The embodiment of the present invention provides a management method for dynamically optimizing the distribution of tourist flow in a scenic area to solve the problems of insufficient dynamic optimization, weak intelligent intervention capability and low overall control efficiency in the existing technology for tourist flow management in scenic areas.

[0007] An embodiment of the present invention provides a management method for dynamically optimizing the distribution of crowds in a scenic area, the method comprising the following steps: In the process of real-time monitoring of tourist flow in the scenic area, tourist density data at the current sampling moment and each previous historical sampling moment are collected to obtain a tourist density data sequence; Extracting local peak and local valley data from the tourist density data sequence, combining the tourist density data between adjacent local peaks and local valleys into a tourist density subsequence, obtaining multiple tourist density subsequences, and calculating the fluctuation characteristic value of each tourist density subsequence; Target tourist density subsequences are selected based on their temporal position. The flow deviation degree of each target tourist density subsequence is calculated based on the difference in fluctuation characteristic values ​​and time span between each target tourist density subsequence and its previous tourist density subsequence, forming a flow deviation degree sequence. Recursively decompose the flow deviation degree sequence to separate the trend item sequence, obtain the first-order difference sequence of the trend item sequence, traverse the first-order difference sequence using a preset sliding window and step size parameter to obtain multiple target analysis windows, and calculate the flow anomaly index of any area in the scenic area at the current sampling time based on the mean of the difference values ​​in each target analysis window; Obtain the traffic anomaly indicators of all areas at the current sampling time, and combine the traffic anomaly indicators of all areas to make global crowd flow distribution optimization decisions.

[0008] Preferably, the calculation of the fluctuation characteristic value of each tourist density subsequence includes: for any tourist density subsequence, performing a first-order difference operation on the tourist density subsequence to obtain a differential data sequence, calculating the mean and variance of the differential data sequence, and using a logarithmic function with a natural constant as the base to map the normalized value of the variance to the sum of the normalized value of the mean to obtain the fluctuation characteristic value of the tourist density subsequence.

[0009] Preferably, the method of screening the target tourist density subsequence according to the time position of each tourist density subsequence includes: selecting the last m tourist density subsequences as the target tourist density subsequence according to the sequential numbering of each tourist density subsequence on the time axis, where m is a preset positive integer.

[0010] Preferably, the calculating of the flow deviation degree of each target tourist density subsequence includes: for any target tourist density subsequence, selecting any tourist density subsequence before the target tourist density subsequence as a reference subsequence, calculating a ratio of the fluctuation characteristic values ​​between the target tourist density subsequence and the reference subsequence, recorded as a first ratio, calculating a ratio between the average density value of the target tourist density subsequence and the average density value of the reference subsequence, recorded as a second ratio, and obtaining a relative deviation degree of the target tourist density subsequence compared to the reference subsequence by multiplying the first ratio by the second ratio; Calculating the absolute value of the difference between the time span of the target tourist density subsequence and the time span of the reference subsequence, and using the negative value of the absolute value of the difference as the independent variable of a preset power function to obtain the influence weight factor of the reference subsequence, where the time span refers to the duration of the corresponding time period of the tourist density subsequence; Combined with the relative deviation degree of the target tourist density subsequence compared to each reference subsequence and the influence weight factor of each reference subsequence, all relative deviation degrees are weightedly summed and the mean is calculated to obtain the flow deviation degree of the target tourist density subsequence.

[0011] Preferably, the calculation of the flow anomaly index of any area in the scenic area at the current sampling moment includes: for any target analysis window, obtaining the average of the differential values ​​in the target analysis window, matching the flow deviation degree associated with each differential value in the target analysis window in the flow deviation degree sequence, and calculating the average flow deviation degree; Using the mean of the differential values ​​as an independent variable of an exponential function with a natural constant as a base to obtain a corresponding exponential value, and calculating a local flow anomaly characteristic value of the target analysis window by multiplying the exponential value by the degree of deviation of the average flow; Combined with the local flow anomaly characteristic values ​​of each target analysis window, the flow anomaly index of any area in the scenic area at the current sampling time is calculated.

[0012] Preferably, the local flow anomaly characteristic value of each target analysis window is combined to calculate the flow anomaly index of any area in the scenic area at the current sampling moment, including: selecting the minimum value among the local flow anomaly characteristic values ​​of all target analysis windows as the reference value, determining the target analysis window to which the current sampling moment belongs, recorded as the window to be analyzed, calculating the ratio of the mean of the differential values ​​in the window to be analyzed to the last differential value, obtaining the difference ratio at the current sampling moment, calculating the ratio of the reference value to the local flow anomaly characteristic value of the window to be analyzed, recorded as the fourth ratio, normalizing the product of the difference ratio and the fourth ratio, and obtaining the flow anomaly index of any area in the scenic area at the current sampling moment.

[0013] Preferably, the process of making a global optimized decision on crowd flow distribution based on traffic anomaly indicators of all regions includes: If the traffic anomaly index of any area in the scenic area at the current sampling time exceeds the preset anomaly threshold, the area will be marked as a suspected anomaly area, and the number of suspected anomaly areas will be counted. If the number is 0, it is determined that the scenic area flow distribution at the current sampling time is normal; If the number is 1, obtain the set of neighboring areas of the suspected abnormal area, recorded as the neighborhood area set, and calculate the difference between the flow abnormality index of each neighborhood area at the current sampling time and the flow abnormality index at the previous sampling time. If there is at least one positive value, conduct crowd diversion intervention in the suspected abnormal area; If the number is greater than 1, the neighborhood area set of each suspected abnormal area is obtained respectively. If there is at least one positive value in the difference between the flow abnormality index at the current sampling moment and the flow abnormality index at the previous sampling moment in each neighborhood area of ​​any suspected abnormal area, crowd diversion intervention is carried out in the suspected abnormal area.

[0014] The present invention constructs a tourist density subsequence to characterize the periodic characteristics of tourist flow changes through real-time collection and analysis of tourist density data in a scenic area, and quantifies the amplitude of the change by calculating the fluctuation characteristic value of each subsequence. Then, based on the difference in fluctuation characteristics between different subsequences, the overall flow deviation degree of each subsequence compared with historical data is evaluated. On this basis, by decomposing the trend term of the flow deviation degree sequence, the key features in the trend term sequence are extracted, and combined with the sliding window analysis method, the flow anomaly index of each area in the scenic area at the current sampling time is calculated to determine whether there is an anomaly in the distribution of human flow. In addition, by introducing the concept of neighborhood area sets and combining the dynamic changes of flow anomaly indicators between multiple areas, a global optimization decision on the distribution of human flow in the scenic area is achieved. This method not only improves the dynamic optimization capability of human flow distribution in the scenic area, but also enhances the efficiency of human flow regulation during peak periods or emergencies, thereby significantly improving the tourist experience and ensuring the safety and stability of scenic area operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 Schematic diagram of the flow of a management method for dynamically optimizing the distribution of tourist flow in a scenic area according to an embodiment of the present invention.

[0016] Figure 2 Schematic diagram of extracting tourist density data sequence and its local peak and valley values ​​in an embodiment of the present invention.

[0017] Figure 3 Schematic diagram of the logic of the calculation process of the tourist density subsequence fluctuation characteristic value in an embodiment of the present invention.

[0018] Figure 4Schematic diagram of trend item decomposition and sliding window analysis of traffic deviation degree sequence in an embodiment of the present invention. Figure 5 This is a flowchart for calculating traffic anomaly indicators and implementing global optimization decisions in an embodiment of the present invention.

[0019] Figure 6 Schematic diagram of the determination and intervention logic of suspected abnormal areas and neighboring area sets in an embodiment of the present invention. DETAILED DESCRIPTION

[0020] The embodiment of the present invention provides a management method for dynamically optimizing the distribution of tourist flow in scenic spots. Figure 1 To the attached Figure 6 The specific embodiments of the present invention are described in detail. Figure 1 Schematic diagram of the flow of the management method for dynamically optimizing the distribution of tourist flow in scenic spots in an embodiment of the present invention. Figure 2 The process of extracting tourist density data series and its local peaks and valleys is shown in the attached figure. Figure 3 The following is a logical diagram of the calculation process of the fluctuation characteristic value. Figure 4 The diagram of trend item decomposition and sliding window analysis of the flow deviation degree sequence is shown in the appendix. Figure 5 This is a flowchart for calculating traffic anomaly indicators and implementing global optimization decisions. Figure 6 This is a logical diagram for determining and intervening in suspected abnormal areas and their neighboring areas.

[0021] In practical applications, it is first necessary to collect tourist density data in real time through the sensor network installed in the scenic area to form a tourist density data sequence. Figure 2 As shown in the figure, an algorithm extracts local peaks and valleys from the visitor density data sequence. These peaks and valleys are used to divide adjacent visitor density subsequences. Each visitor density subsequence consists of the visitor density data between the local peak and valley, thus dividing the entire visitor density data sequence into multiple visitor density subsequences. This process ensures accurate characterization of the periodic characteristics of visitor flow changes.

[0022] Next, for each tourist density subsequence, calculate its fluctuation characteristic value. Figure 3 The calculation process for the fluctuation eigenvalue is demonstrated. For any visitor density subsequence, a first-order difference operation is first performed on the subsequence to obtain a differential data sequence. The mean and variance of the differential data sequence are then calculated, and the normalized variance is added to the normalized mean. This is then mapped using a logarithmic function with a natural constant as the base, ultimately yielding the fluctuation eigenvalue for the visitor density subsequence. This calculation process quantifies the magnitude of the variation in the visitor density subsequence, laying the foundation for subsequent assessments of traffic deviation.

[0023] After completing the calculation of the fluctuation characteristic value, the last m tourist density subsequences are selected as the target tourist density subsequence according to the order number on the time axis, where m is a preset positive integer. For each target tourist density subsequence, its flow deviation degree is calculated and a flow deviation degree sequence is formed. Figure 4 The generation process of the flow deviation degree sequence is demonstrated. Specifically, for any target tourist density subsequence, any previous tourist density subsequence is selected as a reference subsequence. The fluctuation characteristic value ratio between the target tourist density subsequence and the reference subsequence is calculated as the first ratio, and the average density value ratio between the two is calculated as the second ratio. The relative deviation degree is obtained by multiplying the first ratio and the second ratio. In addition, the absolute value of the difference between the time span of the target tourist density subsequence and the time span of the reference subsequence is calculated, and the negative value of the absolute value of the difference is used as the independent variable of the preset power function to obtain the influence weight factor of the reference subsequence. Combining the relative deviation degrees and their influence weight factors of all reference subsequences, all relative deviation degrees are weighted and summed, and the mean is calculated, finally obtaining the flow deviation degree of the target tourist density subsequence.

[0024] In order to further extract the key features of the flow deviation degree sequence, it is recursively decomposed to separate the trend item sequence. Figure 4 The separation process of the trend item sequence and the application of the sliding window are demonstrated. By performing a first-order difference operation on the trend item sequence, a first-order difference sequence is obtained. Then, the first-order difference sequence is traversed using the preset sliding window and step size parameters to obtain multiple target analysis windows. For each target analysis window, the mean of the differential values ​​within it is obtained, and the flow deviation degree associated with each differential value in the target analysis window is matched in the flow deviation degree sequence to calculate the average flow deviation degree. The mean of the differential values ​​is used as the independent variable of the exponential function with a natural constant as the base to obtain the corresponding exponential value. The local flow anomaly characteristic value of the target analysis window is calculated by multiplying the exponential value and the average flow deviation degree. Combined with the local flow anomaly characteristic value of each target analysis window, the flow anomaly index of any area in the scenic area at the current sampling time is calculated.

[0025] Attachment Figure 5 The calculation process of the flow anomaly index and the implementation process of the global optimization decision are demonstrated. The minimum value of the local flow anomaly characteristic values ​​of all target analysis windows is selected as the baseline value, and the target analysis window to which the current sampling time belongs is determined, which is recorded as the window to be analyzed. The ratio of the mean difference value in the window to be analyzed to the last difference value is calculated to obtain the difference ratio at the current sampling time. At the same time, the ratio of the baseline value to the local flow anomaly characteristic value of the window to be analyzed is calculated as the fourth ratio. The product of the difference ratio and the fourth ratio is normalized to finally obtain the flow anomaly index of any area in the scenic area at the current sampling time.

[0026] After obtaining the traffic anomaly indicators of all areas, these indicators are combined to make global optimization decisions on the distribution of passenger flow. Figure 6 The judgment and intervention logic of suspected abnormal areas and neighboring area sets is demonstrated. If the flow anomaly index of any area in the scenic area at the current sampling moment exceeds the preset anomaly threshold, the area is marked as a suspected abnormal area. The number of suspected abnormal areas is counted. If the number is 0, it is determined that the distribution of pedestrian flow in the scenic area at the current sampling moment is normal. If the number is 1, the set of neighboring areas of the suspected abnormal area is obtained, recorded as the neighboring area set, and the difference between the flow anomaly index of each neighboring area at the current sampling moment and the flow anomaly index at the previous sampling moment is calculated. If there is at least one positive difference, the suspected abnormal area is intervened by crowd diversion. If the number of suspected abnormal areas is greater than 1, the neighboring area set of each suspected abnormal area is obtained separately. If there is at least one positive value in the difference between the flow anomaly index at the current sampling moment and the flow anomaly index at the previous sampling moment for each neighboring area of ​​any suspected abnormal area, the suspected abnormal area is intervened by crowd diversion.

[0027] Through the above steps, the present invention realizes the dynamic monitoring, analysis and optimal regulation of tourist flow in scenic areas. Specifically, through the real-time collection of tourist density data series and the extraction of local peaks and local valleys, a tourist density subsequence is constructed to characterize the periodic characteristics of tourist flow changes. The calculation of the fluctuation characteristic value quantifies the amplitude of the change of the tourist density subsequence, and the generation of the flow deviation degree sequence evaluates the overall flow deviation degree of each subsequence compared with the historical data. Through the trend term decomposition and sliding window analysis of the flow deviation degree sequence, the key features are extracted and the flow anomaly index is calculated. Finally, combined with the judgment and intervention logic of the suspected abnormal area and the set of neighboring areas, a global optimization decision on the distribution of pedestrian flow in the scenic area is realized. This method significantly improves the dynamic optimization capability of the distribution of pedestrian flow in the scenic area, enhances the efficiency of pedestrian flow regulation during peak periods or emergencies, thereby improving the tourist experience and ensuring the safety and stability of scenic area operations.

[0028] In order to better enable relevant personnel in this technical field to fully understand and implement the present invention, the specific implementation principle of the present invention is further supplemented below with reference to a specific application scenario.

[0029] In practical applications, firstly, the sensor network installed in the scenic area collects tourist density data in real time to form a tourist density data sequence. Figure 2As shown in the figure, a sensor network is deployed at key locations throughout the scenic area, such as entrances, exits, popular attractions, and main thoroughfares. These sensors collect information on the density of tourists within the area at regular intervals and upload the data to a central processing system. Subsequently, an algorithm is used to extract local peaks and valleys from the tourist density data sequence. A local peak represents the moment when the tourist density reaches its highest point within a certain time period, while a local valley corresponds to the moment when the tourist density is lowest. By identifying local peaks and valleys, the entire tourist density data sequence can be divided into multiple tourist density subsequences. Each tourist density subsequence consists of the tourist density data between the local peak and the local valley, accurately characterizing the cyclical characteristics of changes in tourist flow. This process ensures that subsequent analysis is based on clear cyclical data, avoiding misjudgments caused by data fluctuations.

[0030] Next, for each tourist density subsequence, calculate its fluctuation characteristic value. Figure 3 As shown in the figure, the calculation process of the fluctuation characteristic value involves performing a first-order difference operation on the tourist density subsequence to obtain a differential data sequence. Then, the mean and variance of the differential data sequence are calculated separately and normalized. After adding the normalized variance to the mean, it is mapped using a logarithmic function with a natural constant as the base, and finally the fluctuation characteristic value is obtained. This calculation method can quantify the amplitude of the change in the tourist density subsequence, reflecting the fluctuation characteristics of tourist flow in the time dimension. For example, during peak periods, the fluctuation characteristic value of the tourist density subsequence is generally higher, while during off-peak periods, it is lower. This quantitative indicator lays the foundation for the subsequent evaluation of flow deviation.

[0031] After completing the calculation of the fluctuation characteristic value, the last m tourist density subsequences are selected as the target tourist density subsequence according to the order number on the time axis, where m is a preset positive integer. For each target tourist density subsequence, its flow deviation degree is calculated and a flow deviation degree sequence is formed. Figure 4As shown, the calculation process for the degree of flow deviation involves a comparative analysis between the target visitor density subsequence and a reference subsequence. Specifically, any visitor density subsequence preceding the target visitor density subsequence is selected as a reference subsequence. The ratio of their fluctuation characteristic values ​​is calculated as the first ratio, and the ratio of their average density values ​​is calculated as the second ratio. The relative deviation degree is obtained by multiplying the first and second ratios. Furthermore, the absolute value of the difference between the time span of the target visitor density subsequence and the time span of the reference subsequence is calculated, and the negative value of this difference is used as the independent variable of a preset power function to obtain the influence weight factor of the reference subsequence. Combining the relative deviation degrees and their influence weight factors of all reference subsequences, the weighted sum of all relative deviation degrees is calculated and the average is calculated, ultimately yielding the flow deviation degree of the target visitor density subsequence. This process comprehensively assesses the overall deviation of current visitor flow compared to historical data, providing a basis for subsequent anomaly detection.

[0032] In order to further extract the key features of the flow deviation degree sequence, it is recursively decomposed to separate the trend item sequence. Figure 4 As shown, by performing a first-order difference operation on the trend item sequence, a first-order difference sequence is obtained. Then, the first-order difference sequence is traversed using the preset sliding window and step size parameters to obtain multiple target analysis windows. For each target analysis window, the mean of the difference values ​​within it is obtained, and the flow deviation degree associated with each difference value in the target analysis window is matched in the flow deviation degree sequence to calculate the average flow deviation degree. The mean of the difference values ​​is used as the independent variable of the exponential function with a natural constant as the base to obtain the corresponding exponential value. The local flow anomaly characteristic value of the target analysis window is calculated by multiplying the exponential value and the average flow deviation degree. Combined with the local flow anomaly characteristic value of each target analysis window, the flow anomaly index of any area in the scenic area at the current sampling time is calculated. This process can effectively capture the abnormal fluctuation characteristics in the flow deviation degree sequence and provide a quantitative basis for global optimization decision-making.

[0033] As attached Figure 5As shown, the calculation process of the flow anomaly index further refines the logic of anomaly detection. The minimum value among the local flow anomaly characteristic values ​​of all target analysis windows is selected as the benchmark value, and the target analysis window to which the current sampling moment belongs is determined, which is recorded as the window to be analyzed. The ratio of the mean of the differential values ​​in the window to be analyzed to the last differential value is calculated to obtain the difference ratio at the current sampling moment, and at the same time, the ratio of the benchmark value to the local flow anomaly characteristic value of the window to be analyzed is calculated as the fourth ratio. The product of the difference ratio and the fourth ratio is normalized, and finally the flow anomaly index of any area in the scenic area at the current sampling moment is obtained. This calculation method can comprehensively consider the difference between the current sampling moment and the historical data, as well as the flow anomaly characteristics between different areas, thereby achieving accurate judgment of abnormal areas.

[0034] After obtaining the traffic anomaly indicators of all areas, we combine these indicators to make global optimization decisions on the distribution of people. Figure 6 As shown, if the flow anomaly index of any area in the scenic area at the current sampling time exceeds the preset anomaly threshold, the area is marked as a suspected anomaly area. The number of suspected anomaly areas is counted. If the number is 0, the crowd distribution in the scenic area at the current sampling time is determined to be normal. If the number is 1, a set of neighboring areas of the suspected anomaly area is obtained, recorded as the neighborhood area set, and the difference between the flow anomaly index of each neighborhood area at the current sampling time and the flow anomaly index at the previous sampling time is calculated. If at least one difference is positive, crowd diversion intervention is performed on the suspected anomaly area. If the number of suspected anomaly areas is greater than 1, a neighborhood area set is obtained for each suspected anomaly area. If there is any suspected anomaly area, each neighborhood area has at least one positive value in the difference between the flow anomaly index at the current sampling time and the flow anomaly index at the previous sampling time, crowd diversion intervention is performed on the suspected anomaly area. This process achieves global optimization and control of crowd distribution in the scenic area by combining dynamic changes between multiple areas.

[0035] In summary, this invention constructs visitor density subsequences through real-time collection of visitor density data and extraction of local peaks and valleys to characterize the cyclical characteristics of visitor flow variations. The calculation of fluctuation characteristic values ​​quantifies the magnitude of variation in visitor density subsequences, while the generation of flow deviation degree sequences assesses the overall degree of flow deviation for each subsequence compared to historical data. Through trend term decomposition and sliding window analysis of the flow deviation degree sequences, key features are extracted and flow anomaly indicators are calculated. Finally, by combining the identification and intervention logic for suspected anomaly areas and neighboring areas, a global optimization decision is achieved for the distribution of visitor flow within a scenic area. The visitor density sequence can also be represented using a red, yellow, and green color scheme. For example, red indicates dense crowds and high visitor density, yellow indicates increasing crowds, and green indicates sparse crowd density and low visitor density sequences. Scenic areas should be careful to avoid red. This provides a simpler and more intuitive representation of the visitor density distribution within a scenic area.

[0036] This method significantly improves the dynamic optimization capability of crowd flow distribution within scenic areas and enhances the efficiency of crowd flow control during peak periods or emergencies, thereby improving the visitor experience and ensuring the safety and stability of scenic area operations.

Claims

1. A management method for dynamically optimizing the distribution of tourist flow in a scenic area, characterized in that: The method comprises the following steps: S1. During the real-time monitoring of tourist flow within the scenic area, collect tourist density data at the current sampling time and each previous historical sampling time to obtain a tourist density data sequence; S2. Extracting local peaks and local valleys from the tourist density data sequence, combining the tourist density data between adjacent local peaks and local valleys into a tourist density subsequence, obtaining multiple tourist density subsequences, and calculating the fluctuation characteristic value of each tourist density subsequence; S3. Filter target tourist density subsequences based on the temporal position of each tourist density subsequence. Calculate the flow deviation degree of each target tourist density subsequence based on the difference in fluctuation characteristic values ​​and time span between each target tourist density subsequence and its previous tourist density subsequences, and form a flow deviation degree sequence. S4. Recursively decompose the flow deviation degree sequence to separate the trend item sequence, obtain the first-order difference sequence of the trend item sequence, traverse the first-order difference sequence using a preset sliding window and step size parameter to obtain multiple target analysis windows, and calculate the flow anomaly index of any area in the scenic area at the current sampling time based on the mean of the difference values ​​within each target analysis window; S5. Obtain the traffic anomaly indicators of all areas at the current sampling time, and make a global crowd flow distribution optimization decision based on the traffic anomaly indicators of all areas.

2. The management method for dynamically optimizing the distribution of tourist flow in scenic spots according to claim 1 is characterized in that: The calculation of the fluctuation characteristic value of each tourist density subsequence includes: For any tourist density subsequence, a first-order difference operation is performed on the tourist density subsequence to obtain a differential data sequence, the mean and variance of the differential data sequence are calculated, and a logarithmic function with a natural constant as the base is used to map the normalized value of the variance to the sum of the normalized value of the mean to obtain the fluctuation characteristic value of the tourist density subsequence.

3. The management method for dynamically optimizing the distribution of tourist flow in scenic spots according to claim 1 is characterized in that: The method of screening the target tourist density subsequence according to the time position of each tourist density subsequence includes: selecting the last m tourist density subsequences as the target tourist density subsequence according to the order of numbering each tourist density subsequence on the time axis, where m is a preset positive integer.

4. The management method for dynamically optimizing the distribution of tourist flow in scenic spots according to claim 1 is characterized in that: The calculation of the flow deviation degree of each target tourist density subsequence includes: For any target tourist density subsequence, select any tourist density subsequence preceding the target tourist density subsequence as a reference subsequence, calculate the ratio of the fluctuation characteristic values ​​between the target tourist density subsequence and the reference subsequence, and record it as a first ratio. Calculate the ratio between the average density value of the target tourist density subsequence and the average density value of the reference subsequence, and record it as a second ratio. Multiply the first ratio by the second ratio to obtain the relative deviation of the target tourist density subsequence from the reference subsequence. Calculating the absolute value of the difference between the time span of the target tourist density subsequence and the time span of the reference subsequence, and using the negative value of the absolute value of the difference as the independent variable of a preset power function to obtain the influence weight factor of the reference subsequence, where the time span refers to the duration of the corresponding time period of the tourist density subsequence; Combined with the relative deviation degree of the target tourist density subsequence compared to each reference subsequence and the influence weight factor of each reference subsequence, all relative deviation degrees are weightedly summed and the mean is calculated to obtain the flow deviation degree of the target tourist density subsequence.

5. The management method for dynamically optimizing the distribution of tourist flow in scenic spots according to claim 1 is characterized in that: The calculation of the traffic anomaly index of any area in the scenic area at the current sampling time includes: For any target analysis window, obtain the mean of the differential values ​​within the target analysis window, match the flow deviation degree associated with each differential value in the target analysis window in the flow deviation degree sequence, and calculate the average flow deviation degree; use the mean of the differential values ​​as the independent variable of an exponential function with a natural constant as the base to obtain a corresponding exponential value, and calculate the local flow anomaly characteristic value of the target analysis window by multiplying the exponential value by the average flow deviation degree; combine the local flow anomaly characteristic values ​​of each target analysis window to calculate the flow anomaly index of any area in the scenic area at the current sampling moment.

6. The management method for dynamically optimizing the distribution of tourist flow in scenic spots according to claim 5 is characterized in that: The calculation of the traffic anomaly index of any area in the scenic area at the current sampling time by combining the local traffic anomaly characteristic value of each target analysis window includes: The minimum value among the local flow anomaly characteristic values ​​of all target analysis windows is selected as the benchmark value, and the target analysis window to which the current sampling moment belongs is determined, recorded as the window to be analyzed, and the ratio of the mean difference value in the window to be analyzed to the last difference value is calculated to obtain the difference ratio at the current sampling moment, and the ratio of the benchmark value to the local flow anomaly characteristic value of the window to be analyzed is calculated and recorded as the fourth ratio. The product of the difference ratio and the fourth ratio is normalized to obtain the flow anomaly index of any area in the scenic area at the current sampling moment.

7. The management method for dynamically optimizing the distribution of tourist flow in scenic spots according to claim 1 is characterized in that: The above-mentioned global optimization decision of crowd flow distribution is made by combining traffic anomaly indicators of all areas, including: If the flow anomaly index of any area in the scenic area at the current sampling moment exceeds the preset anomaly threshold, the area will be marked as a suspected abnormal area, and the number of suspected abnormal areas will be counted. If the number is 0, it is determined that the distribution of pedestrian flow in the scenic area at the current sampling moment is normal; if the number is 1, the set of neighboring areas of the suspected abnormal area is obtained, recorded as the neighborhood area set, and the difference between the flow anomaly index of each neighborhood area at the current sampling moment and the flow anomaly index at the previous sampling moment is calculated. If there is at least one positive value, the suspected abnormal area will be intervened to guide the flow of people; if the number is greater than 1, the neighborhood area set of each suspected abnormal area is obtained separately. If there is at least one positive value in the difference between the flow anomaly index at the current sampling moment and the flow anomaly index at the previous sampling moment for each neighborhood area of ​​any suspected abnormal area, the suspected abnormal area will be intervened to guide the flow of people.

8. The management method for dynamically optimizing the distribution of tourist flow in scenic spots according to claim 1 is characterized in that: The length and step size parameters of the sliding window are both preset fixed values.

9. The management method for dynamically optimizing the distribution of tourist flow in scenic spots according to claim 1, characterized in that: The tourist density sequence can also be expressed in the form of red, yellow and green. For example, red indicates dense flow of people and high tourist density, yellow represents increasing flow of people, and green represents sparse flow of people and low tourist density sequence. Scenic spots should pay attention to avoid becoming red. This can more simply and intuitively show the tourist density distribution of the scenic spot.

Images